The success of future fusion power plants as a sustainable energy source greatly depends on their uptime. This uptime relies on the plant's maintenance, which must be performed via tele-manipulation. Tele-manipulated maintenance is challenging, as exemplified by strictly selected and highly trained operators who still require more time to work tele-manipulated, than they do for working hands-on. Many future maintenance tasks involve delicate components that require accurate placement with a dexterous tele-operated slave. Some components are very heavy and need simultaneous hoisting support from a crane, thereby confronting operators with two asymmetric subtasks that have an interactive nature. Literature indicates that having two asymmetric subtasks complicates task execution even more than the already challenging tele-manipulation with one slave system, presumably due to problems in the coordination of the subtasks. Such tele-manipulation with asymmetric slaves must be improved to ensure high plant uptime for future fusion plants. The standard industrial approach to coordinate the control of two asymmetric subtasks prescribes two co-operating operators. However, a single individual could perform the task as well with a bi-manual or hybrid uni-manual control interface. The impact of such differences in control interface design for asymmetric slaves is still a matter of scientific debate. Regardless, the tele-manipulated task will be challenging, and even highly trained operators might benefit from a system that supports them in the task. Although haptic assistance generally improves operator task performance, the main underlying assumption is the availability of perfect knowledge of the task and environment. Handling heavy loads causes manipulator links to deflect statically or dynamically due to their compliance, which cannot be measured or model in sufficient detail. This results in static or dynamic mismatch (inaccuracy) between the real and modelled world that will manifest in the haptic assistive cues, which could negatively affect operator control behaviour. The goal of this thesis is to quantify the impact of interface design choices and haptic assistance to facilitate action coordination between the asymmetric subtasks. Specifically, the interface design choices for single and dual operators will be evaluated with and without haptic assistance, under realistic conditions that incorporate potential inaccuracies in the assistance arising from mismatches between the real and modelled world. The conclusions are that the interface with two co-operating operator is favourable over the bi- and uni-manual single operator interface to coordinate two asymmetric subtasks that have an interactive nature. A novel haptic assistance system improves both co-operated and uni-manual task performance. Interestingly, the observed favour for the co-operated interface with respect to the uni-manual interface is not found when both are haptically assisted. Moreover, haptic assistance still provides benefits when the support cues become statically or dynamically inaccurate due to heavy load handling.

Tele-manipulation of heavy loads typically requires the simultaneous use of two asymmetric slaves: a crane for vertical weight support; and a robot for accurate lateral positioning. The industrial standard prescribes a pair of operators for such tasks (one operator to control each slave), although in principle one operator might control both slaves with a single, hybrid interface. Accurate and safe co-operative handling of the expensive and fragile heavy components is difficult, presumably due to problems in the coordination of the subtasks and the lack of mutual awareness between the two operators. This study proposes a novel haptic assistance system to improve subtask coordination and task performance. Its novelty consists of haptically linking operators/interfaces through the joint task environment. The system's efficacy is evaluated with fifteen pairs of co-operators and fifteen individual uni-manual operators who manoeuvred a heavy load through a bounded path in Virtual Reality. Haptic assistance improves task completion time for both groups. It also reduces control activity and self-reported workload without affecting a number of critical errors made by the operators. Moreover, without haptic assistance, uni-manual operators perform worse than co-operators, but this difference between the interfaces disappears with haptic assistance.

Certain tele-manipulation tasks require manipulation by two asymmetric slaves, for example a crane for hoisting and a dexterous robotic arm for fine manipulation. It is unclear how to best design human-in-the-loop control over two asymmetric slaves. The goal of this paper is to quantitatively compare the standard approach of two co-operating operators that each control a single subtask, to a single operator performing bimanual control over the two subtasks, and a uni-manual control approach.